Issues with vertical Japanese text

[ghost]

sorry if its stupid, I read the manual or doc whatever you call but I could not decide should I use split text block or not, im using dpscreenocr for manga reading, ty.

[danpla]

Thanks for your question. Answering it directly: as far as I can tell, it should not make a difference with vertical Japanese text.

However, thanks to your question, I found some issues in the current version of dpScreenOCR (1.3.0) regarding vertical Japanese text (jpn_vert and potentially other languages that support vertical text, like chi_sim_vert and kor_vert). I will fix them in the next release, but for now the situation is this:

• On Linux, dpScreenOCR can only recognize vertical Japanese text if "Split text blocks" is enabled.
• On Windows, recognition of vertical Japanese text is unfortunately completely broken in the current version, regardless of the "Split text blocks" option.

So, if you are using Windows, you can either switch to another OCR program (e.g. Capture2Text) or wait for the next version of dpScreenOCR.

[danpla]

The fix was made both on dpScreenOCR (74227e14405697897f19ddf1eb24b3c7cdc310ad) and Tesseract (tesseract-ocr/tesseract/issues/3997) sides, and will be available in the next dpScreenOCR version.